A 3D Reconstruction System For Indoor Environment Based On Laser Scanning

Accurate 3D model reconstruction of indoor environment has been a key role in the fields of cultural relics protection, public services, rescue, etc. 3D laser scanning system is widely used in the model reconstruction of indoor environment because of its advantages in measuring distance and high accuracy. High cost of data acquisition and complexity of post processing are the main problems in the process of the 3D model reconstruction of indoor environment based on laser scanning. The goal of this project is to design a low-cost, portable, measuring fast 3D laser scanning system and study the algorithms in 3D point cloud model reconstruction in order to solve the problem of low efficiency of manual data processing.Through the research of 3D laser scanning technology, this dissertation presents a design scheme of a 3D laser scanning system. The whole system is divided into two parts: the 3D laser measurement unit and the interactive module. The spatial 3D information is obtained by the motor control and the laser pulse measuring method in the 3D laser measurement unit. The interactive module is a software based on Qt framework which is used for point cloud display, data preservation, etc. The measurement system we have designed has low cost, measurement fast, easy-to-portable features. The stability and reliability of the whole system have been verified by multiple measurements and experimental analysis, and the system fully meets the requirements of 3D laser point cloud collection.Point cloud registration is the key of the algorithms in 3D point cloud model reconstruction, which is a process of integrating partial datasets acquired from different views into a consistent global model, then model fitting or other further processing can be done. In this dissertation, an algorithm of 3D point cloud registration based on spherical targets has been designed and the algorithms of preprocessing and feature extraction in point cloud processing have also been introduced, finally a reconstruction scheme for indoor environment 3D point cloud model is presented. The feasibility of the scheme and the effectiveness of the algorithm have been showed by the experimental results. The efficiency of data processing has been significantly improved, and the reconstructed 3D point cloud model has achieved the expected results, which has a certain practical significance.